Integration of Ethnomathematics and Deep Learning in Mathematics Education: Cultural and Structural Valuation of Earthquake-Resilient Javanese Traditional Houses

Abstract

This study aims to integrate ethnomathematical values embedded in Javanese traditional architecture with a Deep Learning model to enhance mathematical literacy and disaster awareness. The focus is on Limasan and Limas houses, representing local wisdom and structural resilience against earthquakes. A mixed-method approach with a sequential explanatory design combined field observation, geometric measurement, documentation, and interviews with cultural
figures and educational practitioners. Cultural valuation was analyzed
using Rosa & Orey's (2011) framework, covering functional, aesthetic, ecological, spiritual, and educational aspects. The findings reveal that architectural elements such as pyramid-shaped roofs, the ratio of pillar height to building width, mortise–tenon joints without nails, spatial orientation, and the flexibility of local materials embody mathematical concepts of ratio, similarity, transformation, and
proportion. These values were integrated into ethnomathematicsbased student worksheets and combined with a Deep Learning approach. The implementation resulted in a 30,32 % improvement in students' mathematical literacy while strengthening cultural connection and disaster preparedness. This study highlights the potential of integrating Ethnomathematics and Deep Learning to enrich culturally grounded mathematics learning innovations that are contextual and meaningful.